Management of autoimmune retinopathy treated with intravitreal dexamethasone implant

Abstract

Purpose

The purpose of this study is to evaluate clinical outcomes of autoimmune retinopathy (AIR) in the patients treated with intravitreal dexamethasone implant (IDI).

Method

Twenty-one eyes of 11 AIR patients treated with at least 1 injection of IDI were retrospectively reviewed. Clinical outcomes before and after treatment, including best corrected visual acuity (BCVA), optic coherence tomography (OCT), fundus autofluorescence (FAF), full-field electroretinography (ff-ERG), and visual field (VF) at last visit within 6 and/or 12 months, were recorded.

Results

Among all the patients, 3 had cancer-associated retinopathy (CAR) and 8 had non-paraneoplastic-AIR (npAIR) with mean followed up of 8.52 ± 3.03 months (range 4–12 months). All patients achieved improved or stable BCVA within 6 and/or 12 months after the treatment. Cystoid macular edema (CME) in 2 eyes and significant retinal inflammation in 4 eyes were markedly resolved after single injection. Central retinal thickness (CFT) in all eyes without CME, ellipsoid zone (EZ) on OCT in 71.4% of eyes, ERG response in 55% of eyes, and VF in 50% of eyes were stable or improved within 6 months after treatment. At last visit within 12 months, both BCVA and CFT remained stable in the eyes treated with either single or repeated IDI; however, progression of EZ loss and damage of ERG response occurred in some patients with single IDI.

Conclusion

Clinical outcomes, including BCVA and parameters of OCT, ERG, and VF, were stable or improved after IDI in a majority of AIR patients. Local treatment of AIR with IDI was a good option to initiate the management or an alternative for the patients’ refractory to the systemic therapy but with limited side effect.

Supplementary information

The online version contains supplementary material available at 10.1007/s00417-022-05941-x.

Introduction

Autoimmune retinopathies (AIRs) are a group of immune-mediated inflammatory diseases which can lead to total blindness. It is characterized by acute or subacute progressive visual loss, abnormal electroretinography (ERG), visual field deficits, and findings in the presence of circulating anti-retinal antibodies (ARAs) [1]. AIRs can be categorized as para-neoplastic AIR (pAIR), which includes cancer-associated retinopathy (CAR) and melanoma-associated retinopathy (MAR), or non-paraneoplastic AIR (npAIR) in the absence of malignancy. Despite great progress over decades, management of AIR remains a challenge, and currently, there is no established management protocol. Immunosuppression is still considered to be the most sensible strategy and first-line therapy due to presumed autoimmune nature of the disease [2].

Long-term systemic mediation, including use of steroid and other immunomodulatory agents, was initially advocated but with variable effect [2, 3]. Ferreyra et al.’s study [3] in a case series of 30 eyes demonstrated that immunosuppression treatment resulted in improvement in visual acuity (VA), visual field (VF), cystoids macular edema (CME), ERG, and antibody immunoreactivity. However, there are still many cases irresponsive to systemic steroid treatment [2, 3], which are believed to be the most common outcome in clinical practice [4]. Moreover, long-term use of corticosteroids may cause severe adverse effects such as osteoporosis, hyperglycemia, hypertension, etc. Therefore, local treatment becomes an attractive alternative option for such refractory cases.

Local application of steroid in the AIR aims to either help confirm diagnosis or initiate short-term treatment trial before starting long-term immunosuppressive therapy with less toxic complications. Triamcinolone acetonide (TA) is the most widely used medication in local injection. A case report of CAR described significant increase in best corrected visual acuity (BCVA), complete resolution of CME, and restoration of retinal anatomy on OCT through 2.5 years of 4 times intravitreal injection of TA (IVTA) [5]. Before the procedure, this patient had received intravenous and oral steroid but experienced VA deterioration over the course of dose tapering [5]. Nowadays, intravitreal dexamethasone(DEX) implant (IDI) tends to replace TA to treat non-infectious uveitis or macular edema-associated retinal vascular diseases due to its long-lasting effect and moderate complications that can be easier controlled. DEX implant is a biodegradable copolymer which is able to be continuously released and take effect up to 6 months after intravitreal injection [6].

Reports of treatment of AIR with IDI are scarce. Recently, Abou et al. [7] revealed significant improvement in VA and ERG 2 months after IDI in a 63-year-old male with MAR. Another CAR case of 65-year-old female showed significant improvement in VF and ERG 8 weeks after IDI [8]. Since it is insufficient to reach a valid conclusion from the case report, we performed the study aiming to vindicate the efficacy of IDI in the treatment of both CAR and np-AIR in a larger, consecutive case of patients.

Patients and methods

The Chinese patients with p-AIR or np-AIR who received IDI in the retinal clinics of Beijing Tongren Hospital from February 2020 to March 2022 were retrospectively reviewed. The study was approved by the Ethics Committee of Beijing Tongren Hospital and adhered to the tenets of the Declaration of Helsinki. Informed consent was obtained from all the patients included in the study.

The inclusive criteria were as follows: (1) The patients were selected based on an expert consensus on the diagnosis and management of np-AIR [2]. The essential diagnostic criteria included no apparent cause responsible for visual function abnormality; ERG abnormality (with or without visual field abnormality); the absence of fundus lesions and retinal degeneration or dystrophy that may explain visual function loss; the absence of overt intraocular inflammation; and the presence of serum anti-retinal antibodies. Supportive diagnostic criteria included symptoms with photopsias or scotomas or dyschromatopsia or nyctalopia or photophobia, systemic autoimmune disease (personal or family history), and rapidity of onset of vision change. A workup for systemic tumor was done for all patients by internal medicine physicians. (2) The patients had failed prior treatment with systemic immunosuppressive therapy or effect of IVTA waned before initiation of IDI. All eyes had one or two injections and were followed up at least 4 months. The patients with systemically concomitant use of high doses of steroid and other immunomodulatory agents were excluded.

All patients received a full comprehensive ophthalmic exam at the initial visit. The results that recorded before and after treatment at each follow-up visit included decimal BCVA (converted into logMAR form for statistical analysis), spectral-domain optic coherence tomography (SD-OCT) (Spectralis HRA + OCT, Heidelberg Inc., Germany), fundus autofluorescence (FAF), full-field electroretinography (ff-ERG) (GT-2008 V, Gotech, Chongqing, China), Humphrey visual field (Carl Zeiss Meditec, HFAII 750), and serum ARAs test by Western blotting (only performed before treatment for diagnosis). VF and ERG were not compulsory follow-up tests, especially for those who were presented with advanced disease and sparing function only in small amount of retina centrally. They had severely damaged ERG (e.g., flat response) and VF at baseline as well as OCT with only a small amount of intact EZ. It is difficult to judge progression in these cases. Identification of presence of serum ARAs included total band number and three subtypes with reported pathogenic effects on the retina, including antibodies against recoverin, α-enolase, and carbonic anhydrase II (CA-II).

Treatment protocol

First, remove the tumor burden in all p-AIR patients by operation, radiotherapy, or chemotherapy. Next, the patients received intravitreal injection of 7 mg DEX implant. The patients were required to be followed up every 1 (for the patients with CME or fast progression) or 3 months and received repeated injection every 6 months or sooner if the patients had worsening of symptoms or recurrence of macular edema or active inflammatory infiltration.

Outcome measures

Primary outcome measures included BCVA changes or rate of BCVA change, as well as changes of visual symptoms, FAF, parameters of OCT, ff-ERG, and VF from baseline to the last visit within 6 months from initiation of IDI treatment. Some of above outcome measures within 12 months were also observed but considered an additional data.

Rate of vision changes was calculated by dividing the changes in VA over corresponding number of months. The rate of VA changes before treatment was defined as rate of VA change from the time of onset of symptom to the time of IDI initiation. Improvement/deterioration of BCVA was defined as a 25% gain/decline compared to baseline. Two parameters of SD-OCT, including central foveal thickness (CFT) and horizontal extent for ellipsoid zone (EZ) in fovea scan (6-mm-wide area centered on fovea), were measured to quantitatively assess the severity of retinal anatomic damage. The change of condition on FAF was shown by the change of lesion area of hyper-fluorescence. To assess response of ff-ERG, implicit time and amplitude of a and b wave in dark-adapted 3.0 Max (combined cone and rod response) were recorded. The significant change was defined as more than 40% increase or decrease of any parameter in ff-ERG [9]. VF was performed on the patients with BCVA ≥ 0.05. Mean deviation (MD) was used to quantitatively evaluate the change of VF. Over the course of treatment, improvement or worsening on the VF was defined as a 2 decibel (dB) increase or decrease, respectively, in the MD if the lens, cornea, and vitreous status were stable [9].

Statistical analysis

All statistical analyses were performed with SPSS 19.0 statistics (IBM SPSS Statistics for Windows, NY, USA). The changes in BCVA and parameters of OCT between baseline and follow-up were evaluated by pair t test and or related-samples Wilcoxon signed rank test at a two-side significance level of 5%. There was no selection bias, and all patients treated and followed up were included in the analysis.

Results

Patients’ characteristics and data selection

Clinical characteristics regarding IDI and presence of ARAs for each patient was summarized in Table 1. Twenty-one eyes of 11 patients (45.5% male, 54.5% female) who fit the criteria [2] were evaluated. In this study, patients with np-AIR accounted for 72.7% (8/11), and the average age at onset of the symptom for all patients was 49.0 ± 12.0 years (range 27–69 years). Notably, patient 2 with np-AIR had Sjogren’s syndrome and patient 11 developed CAR before diagnosis of Lung cancer. Each eye received at least one injection, among which 7 eyes received twice with the mean interval of 6 months. The mean follow-up period from initial IDI was 8.52 ± 3.03 months (range 4–12 months). Some patients (patient 1, 3, 4, and 9) were followed up irregularly due to COVID-19 lockdown, but the outcomes still have clinical significance.

Table 1.

Summary of clinical data regarding IDI and presence of anti-retinal antibody for each patient

Patient no./sex/aage, y	Diagnosis	Disease course, y	Presence of ARAs (no. of bands; 3 subtypes)	No. of injection (eye)	Last visit within 6 m/12 m from initiation of IDI (eye)	Medicine before the treatment	Medicine with IDI	Improvement
1/F/51	npAIR	4 y	4; α-enolase( +)	2(OD)	2/12 (OD)	P, C, STA, IVTA	C*	BCVA, CME
2/F/52	npAIR	10 y	8; α-enolase( +)	1 (OU)	3/12 (OD) 6 (OS)	P, C, STA, IVTA	C*	BCVA& EZ(OD) ERG ( OS)
3/M/44	npAIR	1 y	8; α-enolase( +);CA-II( +)	1 (OU)	6 (OU)	IVTA (OU)		No
4/60/F	CAR (endom- -etrial cancer)	6 m	3; α-enolase( +)	1 (OU)	4(OU)	P		Inflammation (OU)
5/M/27	npAIR	10 y	8; α-enolase( +);CA-II( +)	2 (OD) 1 (OS)	3/12 (OD) 3/12(OS)	P, STA (OU), IVTA (OS)		No
6/M/60	npAIR	3 y	6; CA-II( +)	2(OU)	3/12 (OU)	P, C, A, IVTA (OU)	C*	No
7/M/42	npAIR	2 y	2; CA-II( +)	1 (OU)	6 (OU)	None		ERG ( OD)
8/M/36	CAR (colon cancer)	3 y	4; CA-II( +)	1 (OU)	4/8 (OU)	None		No
9/F/46	npAIR	2 y	7; α-enolase( +)	1 (OD) 2 (OS)	1/8 (OD) 5/8(OS)	P		CME (OD) ERG ( OU)
10/F/69	npAIR	3 y	9; α-enolase( +);CA-II ( +)	2 (OD) 1 (OS)	3/8 (OD) 5 (OS)	P		No
11/F/64	CAR (lung cancer)	4 m	3; recoverin( +); α-enolase( +)	2 (OD) 1 (OS)	6/12(OD) 5/12 (OS)	P		BCVA (OU) inflammation (OU)

A, adalimumab subcutaneous injection; C*, oral cyclosporine, 50 mg BID; CA-II, carbonic anhydrase II; m, month; P, oral prednisone; ELM, external limiting membrane; ERG, electroretinogram; FAF, fundus autofluorescence; npAIR, non-paraneoplastic autoimmune retinopathy; CAR, cancer-associated retinopathy; IDI, intravitreal dexamethasone implant injection; IVTA, 4 mg, intravitreal triamcinolone injection; STA, subtenon triamcinolone injection; y, year

All eyes experienced visual deterioration before initiation of IDI. Prior to the injection, 9 patients had been treated with systemic or local immunosuppressive agents, including steroid or/and cyclosporine. The other two patients are treatment-naïve. Among 9 patients treated before, five patients stopped the medication at least 3 months before initiation of the procedure, while 3 patients received IDI concomitant with oral cyclosporine in minimum maintenance doses (Table 1).

Treatment outcomes

The clinical outcomes before and after IDI over the first 6 months were shown in Table 2.

Table 2.

Clinical outcomes before and after IDI treatment

Parameters (mean ± SD)	Baseline				Last visit within 6 months from initiation of treatment
					Results	P value
BCVA (logMAR)	Mean value		1.19 ± 0.70		1.15 ± 0.76 Stable (17/21, 81.0%) Improved (4/21, 19.0%)	0.482 t
	Rate of BCVA change (unit/month)		0.023 ± 0.06		− 0.007 ± 0.06	0.246*
OCT	CFT (um)	With CME (3/21,14.3%)	335.33 ± 244.80		168.0 ± 22.11	0.109*
		Without CME (18/21, 85.7%)	212.33 ± 26.63		207.67 ± 56.99	0.586*
	EZ (um)	Intact (1/21, 4.8%)	3302.52 ± 2333.78		3244.81 ± 2306.89 Stable (15/21, 71.4%) Improved (1/21, 4.8%) Worsened (5/21, 23.8%)	0.807
		Disruption (20/21, 95.2%)
	Significant inflammatory infiltration (4/21, 19.0%)				Improved 4/4 (100%)	N/A
FAF	HAF ring (11/21, 52.4%) Diffuse or fleck HAF (10/21, 47.6%)				Stable (8/21, 38.1%) Improved (9/21, 42.9%) Worsened (4/21, 19.0%)	N/A
ffERG	Abnormal response except for extinguishment (16/21, 76.2%)	a wave	Implicit time (ms)	26.06 ± 3.80	Not performed (5/16) Stable (2/11, 18.2%) Improved (4/11, 36.4%) Worsened (5/11, 45.5%)	N/A
			Amplitude (uV)	62.14 ± 31.13
		b wave	Implicit time (ms)	55.64 ± 5.84
			Amplitude (uV)	135.49 ± 59.78
	Extinguished ERG (5/21, 23.8%)				Not performed	N/A
VF(11 eyes**, MD)	20.71 ± 10.85 dB				Not performed (5/11) Stable (3/6, 50%) Worsened (3/6, 50%)	N/A
Visual symptom	Visual loss (19/21, 90.5%), photophobia (16/21, 76.2%), photopsia (4/21, 19.0%), nyctalopia (3/21, 14.3%), impaired color vision (8/21, 38.1%)				Stable (11/21, 52.4%) Improved (9/21, 42.9%) Worsened (1/21, 4.8%)	N/A

^tPaired samples T test; *Wilcoxon matched-pair signed rank; BCVA, best corrected visual acuity; CFT, central foveal thickness; CME, cystoid macular edema; DA, dark-adapted; EZ, ellipsoid zone; FAF, fundus autofluorescence; ffERG, full-field electroretinography; HAF, hyper-autofluorescent; MD, mean deviation; OCT, optic coherence tomography; VF, visual field; **number of eyes with BCVA ≥ 0.05

Visual symptoms

The most common symptoms of the AIR patients were loss of vision (90.5%) and/or photophobia (76.2%). Others included photopsia, nyctalopia, and impaired color vision. At post-treatment visit, 42.9% patients claimed improvement, 52.4% of patients remained unchanged, and 4.8% complained getting worse.

Visual acuity

In analysis of all eyes, LogMAR BCVA was increased (visual deterioration) prior to IDI initiation despite of systemic immunosuppression therapy. However, at last visit within the first 6 months after treatment, the mean LogMAR BCVA (1.15 ± 0.76) slightly decreased (visual improvement) when compared to the mean baseline logMAR BCVA (1.19 ± 0.70) although it showed no significant difference (P = 0.482). According to the definition described above, 81% of eyes remained stable and 19% eyes showed significant improvement in BCVA. Visual loss did not occur in any of treated eyes.

SD-OCT

Damage of outer retina was observed in 95.2% (20/21) of eyes at the baseline. The most common findings were disruption of EZ and external limiting membrane (ELM) (100%, 20/20). 85% of them (17/20) were combined with thinning of outer nuclear layer (ONL). However, fovea preservation with intact EZ was still seen in 42.9% (9/21) of the eyes. Besides, CME occurred in 3 out of 15 np-AIR eyes (patients 1 and 9), while significant inflammatory infiltration of retina was found in 4 out of 6 p-AIR eyes (patients 4 and 11). Only one eye appeared insignificant change on OCT.

In the eyes with CME, the mean baseline CFT significantly reduced from 335.33 ± 244.80 um to 168.0 ± 22.11 um at 3 month post-treatment (P = 0.109). Figure 1a–d presented marked resolution of CME in patient 1 treated by repeated IDI over 12-month follow-up. In the eyes without CME, the mean baseline CFT was 212.33 ± 26.63 um and remained stable with 207.67 ± 56.99 um over the first 6 months of treatment (P = 0.586). In fact, no eye showed notable retinal thinning during the whole follow-up period.

Fig. 1.

Efficacy of IDI treatment in the AIR patients with different clinical features. Cystoid macular edema in the patient 1 was rapidly resolved 1 month after initial injection (a and b), but recurred at 6 months (c). After re-treatment, the cysts were completely resolved and the macula was well-maintained until month 12 (c–d). EZ restoration was demonstrated in patient 2 with npAIR over time of single IDI treatment, along with BCVA improvement from 0.03 at baseline to 0.1 at 6 month after injection (e–g), The patient did not receive any treatment thereafter and EZ extent showed progressive reduction accompanied by BCVA decrease from 0.1 to 0.05 at month 12 (double head arrow in the g–h). Significant retinal inflammation was observed in the eye of patient 11 and 4 with CAR (i and k). The deposits and inflammatory cells in the different layer of retinal tissues(arrow) were markedly resolved at 1 month after IDI (j and i). Besides, partial ELM reconstruction was noted in the patient 4 at foci of inflammatory infiltration (arrow in the mac photo)

In all eyes after treatment, integrity of EZ remained stable in 71.4% eyes, worsened in 23.8% eyes, and restored in 4.8% eyes (Table 2). The mean horizontal extent measurement of the EZ was 3302.52 ± 2333.78 μm at the baseline. It reduced to 3244.81 ± 2306.89 μm at last follow-up, with average reduction of 57.71 ± 374.37 μm overtime (P = 0.807). Figure 2a–d revealed reduction of EZ extent in one npAIR eye over time of treatment, corresponding to the shrinking of inner ring of HAF ring (Fig. 2e–f). In contrast, it was noted that 1 eye in patient 2 had persistent EZ reconstruction 3 and 6 months after single injection, along with improvement of BCVA from 0.03 to 0.1 (Fig. 1e–g). However, progressive reduction still occurred without continuing treatment accompanied by BCVA decrease from 0.1 to 0.05 at 12 months in this patient (Fig. 1g–h).

Fig. 2.

Progression of disease condition in spite of IDI treatment. EZ extent in the macula of patient 10 was gradually reduced over 8 months of follow-up period (a–d), which is corresponding to the constriction of the inner boundary of hyperfluorescent (HAF) ring (e and f). Meanwhile, significantly reduced ERG response were also shown (g and h). Arrow head: HAF ring; double head arrow: inner boundary of HAF

Two CAR patients (patient 11 and patient 4) with active vitreoretinal inflammation demonstrated significant response to IDI by showing dramatic decrease of inflammatory cells or deposits and even partial restore of ELM (Fig. 1i–l).

FAF

Half of eyes (11/21) had hyper-autofluorescent (HAF) ring around macula and the residuals demonstrated diffuse or multiple fleck HAF. After treatment, 42.3 %((9/21) eyes were improved by indicating decreased area or intensity of HAF, 38.1% eyes (8/21) had no discernable changes, and 19.0% eyes (4/21) showed expanded HAF. As mentioned above, 2 eyes showed constriction of inner boundary of the HAF ring over 6 months of treatment (Fig. 2e and f), indicating progress of the conditions which was consistent with loss of EZ on OCT [10].

ff-ERG and VF

All eyes were performed ff-ERG at baseline and showed moderate to severe damaged rod or/and cone response, including 5 eyes with extinguished response. The mean implicit time and amplitude of a wave and b wave of the other 16 eyes with measurable response at baseline was shown in Table 2. Eleven of these eyes took ff-ERG test within 6 months post-injection. As a result, based on the definition described in the method section, 4 eyes demonstrated significant increase in amplitude of a or/and b wave, 2 eyes remained stable in all response parameters, and 5 eyes showed deterioration with decreased amplitude or delayed implicit time of a or b wave or both. Figure 2g–h showed damage of ERG response in patient 10, consistent with the anatomic damage reflected on OCT and FAF of this patient (Fig. 2a–f).

In the 11 eyes in which visual field test was performed, tube vision, visual defect continuous with physiological blind spot, central or paracentral scotoma, and other patterns of VF defect were observed with mean MD 20.71±10.85 dB at baseline. Six eyes were followed up by the test within 6 months after initial IDI, in which 50% eyes remained stable while the other half showed deterioration.

The detailed information of ff-ERG and VF was provided in the supplementary materials.

Outcome measures within 12-month preliminary data

Thirteen eyes in 8 patients were followed up after the first 6 months of the treatment. The range of the period was from 8 to 12 months. Out of them, 7 eyes received the second IDI (Table 1), in which the mean logMAR BCVA at last visit was 1.85 ± 0.67, slightly worsened compared to the baseline BCVA but without significant difference (1.40 ± 0.72, P = 0.053). Besides, the overall rate of visual loss was also comparable to that prior to IDI initiation (0.033 ± 0.05 unit/month). Nevertheless, we found 5 of these eyes had no remarkable visual change, whereas 2 eyes showed significant BCVA decline due to development of cataract (patient 1) and secondary optic disc atrophy (patient 11). The mean CFT was 143.0 ± 62.48 um at last visit with no significant thinning compared to the baseline (168.0 ± 22.11 um, P = 0.298). EZ extent remained stable in 85.7% (6/7) eyes except for progressive reduction occurred in 1 eye of patient 10 (Fig. 2a–d).

The rest of 6 eyes without repeated IDI were under observation. At last visit, the mean logMAR BCVA had minor increase but generally remained stable (1.09 ± 0.79 vs 0.97 ± 0.77 at the baseline, P = 0.630). The rate of visual decline was 0.012 ± 0.06 unit/month, slightly slower than that prior to IDI initiation (0.041 ± 0.05 unit/month). CFT at last visit remained stable in these eyes, including the right eye of patient 9 with CME (182.33 ± 62.08 um at last visit vs 192.5 ± 57.73 um at baseline, P = 0.463). However, progressive loss of EZ was found in 50% eyes (3/6), including patient 2 (Fig. 2g–h).

Complication

Drug-associated complications included mild increase of intraocular pressure in 4 eyes and development of posterior subcapsular opacity in lens in another 2 eyes with repeated injection. The IOP increase was well controlled by anti-glaucoma eye drop and one eye needed cataract surgery.

Discussion

In an expert consensus on diagnosis and management of AIR [2], supermajority agrees that more evidence is needed for considering long-term immunomodulatory therapy. Moreover, poor treatment outcomes of the therapy targeting decrease of circulating ARAs [11] weaken the strategy of systemic use of steroid which may cause severe adverse effects. Local intraocular immunosuppression thus provides an alternative therapeutic approach in treatment of the disease. In our case series study of both CAR and np-AIR patients included, IDI achieved favorable results by demonstrating stability or improvement in visual function and retinal anatomy, measured by multimodal imaging in most patients during the follow-up period within 6 months. Furthermore, the patients with CME or significant retinal inflammation displayed high response to IDI treatment. Despite disease progress occurred in some eyes, IDI is still proved to be a good option to initiate the management of the AIR or for the patients refractory to the systemic therapy.

As mentioned above, the AIR patients of our cohort responded well to IDI in two forms of conditions which indicated fast progress of the disease [12]. CME is one of complications in AIR, commonly seen in npAIR [3] with the occurrence from 23% [3] to 75% [12], but rarely in CAR [13] or MAR. The study by Finn et al. [12] showed CME was associated with decreased ff-ERG amplitudes and greater velocity of EZ loss, suggesting that CME is a biomarker of more severe and more progressive disease in npAIR. CME in the AIR was reported to be highly responsive to systemic immunosuppressives combined with local injection of TA or IVTA alone, resulting in improvement in macular anatomy and partial visual function [3, 5, 13]. In this study, we first tested the effect of IDI on the treatment of CME in AIR. Two of three eyes with CME achieved remarkable resolution after single IDI along with significant BCVA improvement in one eye. The effect maintained up to 6 months which was much longer than IVTA, but still needed repeated injection when the treatment wore off.

The second form of disease condition that responded dramatically to IDI was significant vitreoretinal inflammation. AIR is typically presented with minimal or no signs of intraocular inflammation [14, 15]. However, inflammatory response in CAR is more obvious compared to np-AIR. Huynh et al. [5] reported a CAR case showing multiple tiny disruption of the EZ and hyperreflective material, indicative of foci of active inflammation. After treated by IVTA, the inflammatory lesion improved and EZ restored [5]. In our study, 4 eyes in two CAR patients (patient 4 and patient 11) showed marked resolution of inflammatory reaction on OCT and FAF after IDI treatment, along with significant BCVA improvement (patient 11). Moreover, patient 4 showed ELM reconstruction in one eye at foci of severe inflammatory infiltration.

In spite of improvement in some cases, prognosis of AIR was usually poor due to delayed diagnosis and lack of standardized treatment protocols. Experts agreed that early treatment is critical for AIR [2, 16] and the purpose of treatment is largely for stability of the disease. In our study, the disease conditions following treatment were maintained in half or more patients by showing no further vision loss (100%), no worsening on parameters of OCT (76.2%), FAF (81%), ff-ERG (54.6%), VF (50%), and patients’ visual symptoms (95.3%) over 6 months of treatment. These were comparable to the results of systemic therapies described in the literature but with limited side effect [3, 9, 17]. Although it is unknown whether IDI has better clinical outcomes than IVTA in treatment of AIR, the side effect of sustained-release steroid implant has been shown to be milder than the IVTA which may cause uncontrolled intraocular pressure (IOP) increase that sometime needs glaucoma surgery.

Prior reports reveal that despite aggressive immunosuppressive therapy, rapid and relentless visual loss often occurs. The therapy is not helpful once widespread retinal degeneration develops [18]. In our cohort, the patients in the advanced stage (poor VA, tube vision, flat ff-ERG response, or complete loss of EZ on OCT) accounted for more than 50% and 27.3% of them experienced worsening of conditions anatomically and functionally after treatment. All these eyes were npAIR with the disease course ranging from 2 to 4 years. It was noted that progressive reduction of EZ extent on OCT was indicated in these eyes, whereas their CFT did not show significant thinning, suggesting EZ extent was more sensitive than retinal thickness as a biomarker to evaluate the disease progression. EZ restoration in the AIR is rare but crucial in recovery of visual function. In our study, only 1 eye in patient 2 with neither CME nor significant retinal inflammation demonstrated EZ reconstruction along with improvement in BCVA from 0.03 at baseline to 0.1 after treatment. FAF is another noninvasive tool to evaluate AIR. HAF ring was a common sign in AIR [19]. A typical case in our study displayed disease progression by constriction of the inner boundary of HAF ring, which was consistent with loss of EZ on OCT and ff-ERG deterioration. These results suggested that there was still a certain number of cases refractory to local treatment alone. Initial combination or prompt switch to systemic steroid or other treatment modalities are quite essential.

A preliminary analysis of longer-term efficacy of IDI was also performed in this study. According to the outcome measures (particularly BCVA and OCT parameters) at last visit within 12 month, both single and repeated injection were effective in maintaining VA and CFT or slowing the rate of VA loss when compared with before injection. However, repeated injection seemed to play a role in reducing the chance of damage of EZ and ff-ERG response in some patients. Although effect of repeated injection seemed to be inferior to the single injection in terms of overall BCVA improvement at longer time observation, it partly resulted from complications associated with multiple intravitreal use of steroid (e.g., development of cataract) and can be treated by surgery if needed. Nevertheless, close follow-up with larger samples and more comprehensive test data are warranted to explore the long-term efficacy of local therapy in stabilization of disease condition.

The limitation of this work is the nature of retrospective analysis, comparably small number of cases (rare disease), lack of control group, and irregular follow-up visits. Larger case numbers with careful longer-term of follow-up, control groups with blank, or other treatment approaches, e.g., systemic steroid, may be used to further confirm the conclusion or indicate the difference between local and systemic use of steroid.

In summary, our current study demonstrated the safety and efficacy of IDI for stabilization of disease condition and may achieve anatomic and functional improvement particularly in presence of CME and significant inflammatory reaction. The outcome data from our cases series further encourage the physicians in this field to consider sustained-release intraocular steroid implants as a preferred alternative to systemic corticosteroids therapy for AIR, especially for initiation of management.

Supplementary information

Below is the link to the electronic supplementary material.

Funding

This work was founded by Beijing Municipal Natural Science Foundation (7192034, H.Y.Z) and Key research program of Beijing Institute of Ophthalmology (2020, H.Y.Z.)

Declarations

Ethical approval

All procedures performed in studies involving human participants were in accordance with the ethical standards of the Ethics Committee of Beijing Tongren Hospital and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Consent to participate

Informed consent was obtained from all individual participants included in the study.

Conflict of interest

The authors declare no competing interests.